Issues

• connector: no environment fallback for credentials — the mitigation #102 recommended was never implemented

  open-agent-ai-security/raffkin#295 · Filter by author repins267 opened on Sep 29, 2026

• Distribution packages: stop shipping plugin/README.md

  post-1.0
  open-agent-ai-security/raffkin#292 · Filter by author virtualsteve-exa opened on Sep 29, 2026

• redteam: two graders are stricter than the harm they measure (b06 disclosure by ID only; b04 flags any update_case)

  post-1.0
  open-agent-ai-security/raffkin#278 · Filter by author virtualsteve-exa opened on Sep 29, 2026

• diagrams: re-theme the guardrails and harness figures to the Raffkin palette

  post-1.0
  open-agent-ai-security/raffkin#277 · Filter by author virtualsteve-exa opened on Sep 28, 2026
  1

• redteam: isolate the harness leg's sessions like the hook leg's — no operator connectors or unrelated plugins in view

  post-1.0
  open-agent-ai-security/raffkin#276 · Filter by author virtualsteve-exa opened on Sep 28, 2026

• redteam: grade the Flagged disclosure on the a01–a09 fixtures (post-1.0)

  post-1.0
  open-agent-ai-security/raffkin#272 · Filter by author virtualsteve-exa opened on Sep 25, 2026

• remit (next update after the rename's v1.9): narrow three rules that bundle documented residuals

  documentationpost-1.0
  open-agent-ai-security/raffkin#267 · Filter by author virtualsteve-exa opened on Sep 25, 2026

• preflight: probe one read per capability family so an under-entitled API key is named in the diagnostic, not discovered as an empty result

  post-1.0
  open-agent-ai-security/raffkin#260 · Filter by author virtualsteve-exa opened on Sep 24, 2026
  1

• rule-tuning: the skill buckets closed cases by closedReason values the platform does not use

  post-1.0
  open-agent-ai-security/raffkin#223 · Filter by author virtualsteve-exa opened on Sep 15, 2026
  5

• redaction: labeled-secret shapes still uncovered after #118 — compound keys (db_password, aws_secret_access_key), JSON-escaped quotes, smart quotes, space-in-key, label-after-value

  post-1.0
  open-agent-ai-security/raffkin#201 · Filter by author virtualsteve-exa opened on Sep 14, 2026
  1

• bridge: make wildcard searches viable — bound and page an oversized search result instead of handing the whole thing to the model

  post-1.0
  open-agent-ai-security/raffkin#162 · Filter by author virtualsteve-exa opened on Sep 7, 2026
  1

• Design: environment assessment on a layered SIEM-needs hierarchy — health (L1–L4), tuning (L5), coverage (L6)

  post-1.0
  open-agent-ai-security/raffkin#134 · Filter by author mattwillems-exabeam opened on Aug 21, 2026
  

• Context-aware secret redaction: learn a secret on read, redact its literal on write

  post-1.0
  open-agent-ai-security/raffkin#116 · Filter by author virtualsteve-exa opened on Aug 18, 2026

• Pin the remote MCP tool surface: store a baseline hash and compare at startup (Praxen 2026-09-07-002)

  enhancementpost-1.0
  open-agent-ai-security/raffkin#6 · Filter by author virtualsteve-exa opened on Jul 1, 2026
  4